HYPE Market Intelligence — Week of September 21–27, 2026
The week in one line: HYPE set a fifth all-time high at $98.00, closed $91.72 — down 2.2% — and did it in a week when both majors rose, breaking a five-week run of outperformance. This report works through the price ladder, the valuation arithmetic (with a fee window that is honestly incomplete), the buyback capacity ledger, relative strength on identical clocks, and the five regime changers published here last week, graded after the fact.
1. The price ladder
Five records in five weeks, and the fifth arrived on the shortest interval of the cycle:
| Record | Date | Intraday high | Close that day | Interval since prior record |
|---|---|---|---|---|
| 1st of the cycle | Aug 23 | $83.46 | $82.37 | First close above the May high of $67 |
| 2nd | Aug 27 | $86.73 | — | 4 days |
| 3rd | Sep 6 | $89.60 | $87.94 | 10 days |
| 4th | Sep 19 | $94.60 | $92.22 | 13 days |
| 5th | Sep 23 | $98.00 | $94.09 | 4 days |
The path between the fourth and fifth records is the whole story of this week. The Sep 19 record came 13 days after the prior one, then the fifth fell just four days later — a compression of the interval that was not sustained. The Sep 23 session printed $98.00 and closed $94.09, a 4.0% reversal from its own high, and the token never traded near the peak again: the week closed $91.72, which is 6.4% below the record and 2.2% below the prior Sunday's close of $93.75.
One measurement discipline note, because it matters for anyone comparing coverage: the record is venue-specific, but this time the venues agreed far more closely. Hyperliquid's own candle series printed $98.00; same-day coverage carried $97.96 and $97.95; and a market column filed during the Sep 22 session recorded $96.68 for a print that occurred earlier in the run. The spread between the highest and lowest record figures is about 1.4%, against 4.0% for the September 19 record — because this peak happened in a single liquid session rather than across two thin ones. The record still belongs to a venue; this week it just happens not to matter much which one you quote.
2. Valuation arithmetic on verified inputs
From verified inputs only, with one honest caveat stated up front: the fee window is now the full seven days: DefiLlama published the Sep 27 print ($1.75M) after this report’s Sep 28 fetch. Market cap at the weekly close ($91.72 × 222.446M circulating) is $20.40B. The seven-day fee total is $19.46M, which annualised at 52 weeks is a $1.01B run-rate; the 30-day base ($69.3M × 52 ÷ 4.33) annualises to $0.83B. On fully diluted supply of 955.3M the same market value is $87.6B.
| Inputs | Value | Multiple on circulating | Multiple on FDV |
|---|---|---|---|
| Market cap / 7-day fee run-rate ($1.01B) | $20.40B | ~20.2x | ~87x |
| Market cap / W5 7-day ($19.34M) | $20.40B | ~20.3x | ~87x |
| Market cap / 30d fee base ($0.83B) | $20.40B | ~24.6x | ~106x |
| Fully diluted value / 7-day fee run-rate ($1.01B) | $87.6B | ~87x | n/a |
Read the first two rows together, because they bracket the honest answer. On the full seven prints, the multiple is ~20.2x — essentially unchanged from last week’s ~20.3x (W5, also seven prints). The 22.2x figure shown at fetch time was a missing-print artifact (six of seven days on the numerator); the resolved 7-day multiple is marginally tighter than last week, not wider. The defensible statement is therefore not "HYPE de-rated" or "HYPE re-rated" but "the multiple is between roughly 19x and 22x this week, and the spread is the data gap, not the market". The bottom row exists to make the same point from the other side: the identical $20.40B market cap produces 19.0x to 24.6x depending entirely on which fee window you annualise. These are restatements, not targets.
3. The buyback ledger, stated carefully
What is verifiable: gross protocol fees of $19.46M over seven days and $69.3M over 30 days, and the documented mechanism that the overwhelming majority of protocol revenue routes to the Assistance Fund for HYPE buyback-burn, supplemented by the USDC aligned-quote-asset reserve-yield share. What is not verifiable weekly: the AF's actual USD purchases — the ledger is on-chain but has no clean API, so this report derives capacity rather than claiming executed amounts.
| Line | Value | Basis |
|---|---|---|
| Gross fees, 7 days | $19.46M | DefiLlama dailyFees, summed (Sep 27 $1.75M published post-fetch) |
| Implied weekly buyback capacity | ~$19.3M | ~99% pass-through (third-party estimate) × fees |
| Implied capacity at prior week's fees | ~$19.1M | Same construction, $19.34M week (resolved) |
| Cumulative HYPE burned | ~48.7M (~4.9% of max supply) | Press tally; no single public ledger API |
| 2026 year-to-date revenue | ~$429M, first among crypto projects | Press tally; narrower measure than gross fees |
| Reserve-yield contribution | ~$135M–$182M/yr estimated | Third-party estimate; first payment scheduled Oct 3 |
| Weekly executed USD purchases | Unavailable | No public API series |
Read the second and third rows together: capacity held essentially flat week over week (+1%), tracking the now-resolved fee total ($19.3M vs W5’s $19.1M). That is the mechanical consequence of near-total pass-through, and it is the least flattering true thing to say about the buyback: the programme does not smooth revenue, it transmits it. In a week when fees decayed from $4.65M on Monday to $1.17M on Saturday, the implied buy bid decayed with them. October 3 is the date that converts the largest modeled line in this table into a measured one, and it is the single most informative scheduled event on the calendar after the September 29 unlock.
4. Relative strength, same-source
| Week | HYPE | BTC | ETH |
|---|---|---|---|
| Aug 24–30 | −2.9% | −0.2% | −1.9% |
| Aug 31–Sep 6 | +9.9% | +3.4% | +4.0% |
| Sep 7–13 | −12.1% | −4.4% | −1.6% |
| Sep 14–20 | +21.3% | +5.7% | +6.9% |
| Sep 21–27 | −2.2% | +4.0% | +1.6% |
| Five-week cumulative (Aug 23 → Sep 27) | +11.4% | +8.5% | +9.1% |
The same API, the same clocks — every cell is Sunday close over Sunday close, so the comparison is clean. One reading matters, and it is new. For four weeks the pattern was amplitude: HYPE moved roughly three times the majors in whichever direction the week took, and the cumulative column carried a large excess (64.2% versus 29.2% a week ago). This week the sign flipped without the amplitude — HYPE fell 2.2% while BTC rose 4.0% and ETH rose 1.6% — and the rerun of the cumulative table tells you why that matters more than the weekly number: with the late-August breakout week now outside the window, five-week HYPE performance is +11.4% against BTC's +8.5%. The excess is still positive but it has shrunk from 35 points to under 3, and the reason is not that HYPE fell apart; it is that the window that made the outperformance look structural has rolled off. Any claim about a token's relative strength is a claim about a start date, and moving the start date by one week changes this one by 30 points.
5. What would change the regime
Five regime changers were published here last week. Grading them is the most useful thing this section can do:
| Regime changer (set Sep 21) | What happened | Status |
|---|---|---|
| "A fee print above $16.5M would show the contraction was volatility-driven, not demand-driven" | Seven-day fees printed $19.46M — above the line | Triggered — but the week's internal decay ($4.65M → $1.17M) points the other way |
| "The September 29 unlock (14,175,778 HYPE, ~$1.33B at the prior week's close)" | Still ahead; notional now ~$1.2B at this week's prices | Unresolved |
| "Funding above 0.05%/8h while price stalls — the earliest crowding warning" | Funding fell to ~0.000%/8h while price stalled | Did not trigger — crowding unwound rather than built |
| "A disclosed AQAv2 payment on October 3" | Still ahead | Unresolved |
| "Lending utilization decay — a day-one draw that decays says arbitrage; one that holds says need" | No public API publishes lending utilisation | Unmeasured |
For next week, the changers:
- A full seven-print fee week above $16.5M. The resolved week cleared the line at $19.46M, though on a weekday decay that argues volatility over structure. A complete week above it, with the weekday decay reversed, is the version that would settle the structural question.
- The September 29 unlock (14,175,778 HYPE, ~$1.2B at this week's close). The reference level is now $89.77, the week's low — not the $75.15 line that was never approached. A close below $89.77 into the unlock would confirm the reversal reading; a reclaim of $94.60 would keep consolidation alive.
- $98.00 turning from a print into a reference. A failed record becomes resistance. Two consecutive closes above $94.60 — the prior record — would read as coiling beneath a new high; anything under $89.77 reads as retracement.
- Funding rebuilding from zero. At ~0.000% the base is clean. A move back above 0.05%/8h while price stalls would be a much sharper crowding signal than last week's 0.039% reading ever was, precisely because it would have to be built rather than inherited.
- A disclosed AQAv2 payment on October 3. The only scheduled event that moves a line in this report from "Modeled" to "Confirmed".
No price targets here — the levels above are descriptions of the map, not predictions of the route.
6. The week, day by day
| Date | Close | Intraday range | Δ | Fees (day) |
|---|---|---|---|---|
| Mon Sep 21 | $94.14 | $92.00 – $96.15 | +0.4% | $4.65M |
| Tue Sep 22 | $97.27 | $92.16 – $97.82 | +3.3% | $3.08M |
| Wed Sep 23 | $94.09 | $91.95 – $98.00 | −3.3% | $2.75M |
| Thu Sep 24 | $92.08 | $89.77 – $95.00 | −2.1% | $3.59M |
| Fri Sep 25 | $92.35 | $90.25 – $94.37 | +0.3% | $2.47M |
| Sat Sep 26 | $92.59 | $91.02 – $92.69 | +0.3% | $1.17M |
| Sun Sep 27 | $91.72 | $91.02 – $94.00 | −0.9% | $1.75M |
The intraday column is where the week's structure lives. Wednesday's range — $91.95 to $98.00, with the close at $94.09 — is a 6.6% round trip inside one session that finished 4.0% below its peak. Thursday then took out Wednesday's low and printed the week's $89.77 low, and Friday through Sunday traded a $91–$94 band with progressively thinner fees. That is the anatomy of a distribution: the high is made once, under the loudest conditions, and the rest of the week is the slow transfer of the float that bought the print.
The fee line tells its own version of the same story, and it is the most uncomfortable number in this report. Monday's $4.65M was the largest single-day print since early August, and it landed on a day when HYPE rose just +0.4% while BTC gained 6.7% — the fee cost of positioning into a record rather than the revenue of having one. Tuesday's +3.3% session generated only $3.08M, and by Saturday the line had decayed to $1.17M, the lowest weekday-or-weekend reading since early September. A fee series that peaks on the first day of a record week and bleeds for six sessions is a volatility proxy doing exactly what a volatility proxy does — which is why the honest weekly read is "event-driven", not "expanding".
Frequently asked questions
How can the multiple widen and the token fall in the same week?
Because the numerator and denominator are measured over different windows. Market cap fell 2.2% while the seven-day fee total ($19.46M) annualised to a run-rate near last week's, so the multiple barely moved — ~20.2x versus W5's ~20.3x. The 22.2x figure shown at fetch time was a missing-print artifact (only six of seven days on the numerator); the resolved 7-day multiple is marginally tighter than last week, not wider. The honest answer is a narrow range, and the small spread is the data gap, not a market move.
Which all-time-high figure should I use?
The one belonging to the venue you are quoting — but this week it hardly matters. Hyperliquid's own candles printed $98.00; same-day coverage carried $97.96 and $97.95; and a column filed during the prior session recorded $96.68 for an earlier print. The spread among the record figures is about 1.4%, against 4.0% a week ago, because this peak was made in one liquid session. This report uses the exchange's own series and prints the spread.
Does the buyback capacity figure mean $19.3M was actually spent?
No. It is capacity: gross fees times the estimated ~99% pass-through rate. The AF's executed purchases live on-chain with no clean public API, so this report computes what the mechanism could absorb and labels it as capacity. When a disclosure endpoint exists — the Oct 3 AQAv2 payment is the closest scheduled thing — this table switches from capacity to actuals.
Why does the five-week cumulative look so different from last week?
Because the window moved. Last week's five-week table started on Aug 16 and captured the +44.3% breakout week; this week's starts on Aug 23 and does not. The same token, measured over two windows differing by one week, shows +64.2% versus +11.4%. That is not a contradiction — it is the reason this report always prints its start date next to its cumulative, and the reason any "HYPE has outperformed by X" claim is uninterpretable without one.
Is this financial advice?
No. This is data analysis for informational purposes. Nothing here is a recommendation to buy, sell or hold HYPE or any other asset.